Volcanology, Geochemistry, Petrology [V]

V52B MCC:3007 Friday

The 2004-2005 Eruption of Mount St. Helens I

Presiding: M A Clynne, U.S. Geological Survey; R P Denlinger, U.S. Geological Survey

V52B-01 INVITED

Measuring the Dome Growth of the 2004-2005 Eruption of Mount St. Helens

* Schilling, S P (sschilli@usgs.gov) , US Geological Survey Cascades Volcano Observatory, 1300 SE Cardinal Court Bldg 10 Ste 100, Vancouver, WA 98681 United States
Denlinger, R (roger@usgs.gov) , US Geological Survey Cascades Volcano Observatory, 1300 SE Cardinal Court Bldg 10 Ste 100, Vancouver, WA 98681 United States
Thompson, R A (rathomps@usgs.gov) , US Geological Survey, Denver Federal Center, Denver, CO 80225 United States
Messerich, J (jmesser@usgs.gov) , US Geological Survey, Denver Federal Center, Denver, CO 80225 United States

In October 2004, a new period of dome growth began that changed the topography of the 1980 crater at Mount St. Helens dramatically. From October 2004 through July 2005, nearly 60 million cubic meters of lava extruded onto the crater floor immediately south of the 1980-1986 lava dome. The eruption intensely deformed and divided the crater glacier on this floor. It created spectacular crevassing and rapid advance of the east arm of the glacier then caused crevassing and broad uplift of the glacier's west arm. Time-sequential vertical aerial photography documents morphologic change and enabled construction of a series of 13 2-m-resolution digital elevation models (DEMs) for the period between October 4, 2004 and July 14, 2005. Vertical aerial photographs flown at a nominal 1:12000 scale were acquired at three-week intervals, scanned at 12-micron resolution, and rectified using a soft-copy (i.e., digital image) photogrammetric workstation. Aerotriangulated models were constructed using ground control outside the area of active deformation, derived from pre-eruption GPS and photogrammetric data, and passed to subsequent model sets. Resulting location accuracy is on the order of decimeters. The DEM data allow us to estimate dome volumes during growth. To extract volumetric changes and calculate extrusion rates, each DEM surface was compared to pre-eruption reference surfaces from 2000 and 2003, as well as to the preceding DEM surface. On July 14, 2005, the new dome was approximately 700 m long (NW-SE) and 560 m wide (SW-NE). The volume of the new dome (including talus), was about 58 million cubic meters, approximately two-thirds the volume of the 1980-1986 dome. The volumetric growth rate in 2004-2005 ranged from a maximum of 9 m3/sec in the early stages of growth to an average of 1-3 m3/sec thereafter. The DEMs also are used to quantify dome height variations, size of the conduit opening, and the mechanics of dome emplacement (growth and collapse) as well as deformation of glacial ice. Between February 1 and June 15, 2005, the highest point of the new dome remained between 2330 m and 2340 m, but reached 2367 m on July 14, nearly the height of the lowest point on the crater rim (2370 m). Over time, migration of the extrusion site limits the diameter of the conduit near the surface to 150-230 m in an east west direction, whereas the diameter in the NS direction is unconstrained. The lava emerging from this conduit appears to move unimpeded through the glacier ice on the south crater floor, but is diverted by deposits from previous extrusions. The extrusion axis was initially southward, and shifted toward the east from October 2004 through April 2005 as successive whaleback-shaped monoliths emerged from the vent area, displacing older parts of the growing dome. The axis reverted to south during May-June, then shifted westward during July-August as extruding material sought out paths of lesser resistance.

V52B-02 INVITED

Overview of seismicity associated with the 2004-2005 eruption of Mount St. Helens

* Moran, S C (smoran@usgs.gov) , U.S. Geological Survey, Cascades Volcano Observatory, 1300 SE Cardinal Ct., Vancouver, WA 98683 United States
Malone, S D (steve@ess.washington.edu) , University of Washington, Dept. of Earth and Space Sciences, Box 351310, Seattle, WA 98195 United States
Qamar, A I (tony@ess.washington.edu) , University of Washington, Dept. of Earth and Space Sciences, Box 351310, Seattle, WA 98195 United States
Thelen, W (wethelen@ess.washington.edu) , University of Washington, Dept. of Earth and Space Sciences, Box 351310, Seattle, WA 98195 United States
Waite, G (gwaite@usgs.gov) , U.S. Geological Survey, 345 Middlefield Rd., Menlo Park, CA 94025 United States
Horton, S (horton@ceri.memphis.edu) , University of Memphis - CERI, 3876 Central Ave., Memphis, TN 38152 United States
Lahusen, R G (rlahusen@usgs.gov) , U.S. Geological Survey, Cascades Volcano Observatory, 1300 SE Cardinal Ct., Vancouver, WA 98683 United States
Major, J J (jjmajor@usgs.gov) , U.S. Geological Survey, Cascades Volcano Observatory, 1300 SE Cardinal Ct., Vancouver, WA 98683 United States

Following 18 years of quiescence, an earthquake swarm on September 23, 2004, heralded the reawakening of Mount St. Helens (MSH). Seismicity during the 18-year quiescence consisted of shallow (< 3 km) and deep (3-10 km) volcano-tectonic (VT) earthquakes, with the deeper earthquakes occurring mostly during several extended periods of elevated seismicity that likely reflected intrusions of new magma. The swarm on September 23 intensified in a step-wise manner over the first week, and by the first phreatic explosion on October 1 earthquakes had increased significantly in size (Mmax = 3.5) and rate (1-2 per minute). The earthquakes also became shallower (< 1 km) and transitioned from VT to hybrid in character. Following a final phreatic explosion on October 5 earthquake magnitudes and rates decreased significantly. One of the remarkable aspects of MSH seismicity has been the occurrence of repetitive (in waveform) and regular (in time) shallow "drumbeat"" hybrid earthquakes that began in mid-October as lava dome extrusion commenced. With a few exceptions drumbeats occurred roughly every 40-60 seconds through April 2005 before progressively increasing to over 1000 seconds by mid-August. They occurred in association with extrusion of a fault-gouge-covered lava dome that emerged at a relatively constant rate, a linkage that strongly suggests a stick-slip origin for the drumbeats. The increase in inter-event times does not appear to directly correlate with decreasing extrusion rate, suggesting that dome extrusion has become increasingly aseismic over time. Another remarkable aspect is the occurrence of shallow "big"" earthquakes (M 2.5-3.5) starting in mid-November in association with the breakup of the first erupted dome. To date there have been over 250 of these events that have occurred in four temporally-defined groups, most associated with periods when extruded domes were breaking apart. Roughly half have impulsive arrivals, and many of these have dominantly down first motions consistent either with very shallow reverse faulting or a non-double-couple source (such as failure along a curvilinear fault). Broadband recordings show that very long-period (VLP) events with 10-30s periods accompany many of the ""big"" events as well as many of the drumbeat events. The impulsive big events may reflect large-scale lurches of the new dome, perhaps beginning as slow-slip events as indicated by VLP occurrence. Big events with emergent arrivals have recently been found via time-lapse photography to be associated with slumps of the dome, and thus appear to be directly related to failures within the dome following extrusion.

V52B-03

Ground Deformation Associated with the 2004-2005 Dome-building Eruption of Mount St. Helens, Washington

* Dzurisin, D (dzurisin@usgs.gov) , USGS Cascades Volcano Observatory, 1300 S.E. Cardinal Court Building 10, Suite 100, Vancouver, WA 98683-9589 United States
Lisowski, M (mlisowski@usgs.gov) , USGS Cascades Volcano Observatory, 1300 S.E. Cardinal Court Building 10, Suite 100, Vancouver, WA 98683-9589 United States
Schilling, S P (sschilli@usgs.gov) , USGS Cascades Volcano Observatory, 1300 S.E. Cardinal Court Building 10, Suite 100, Vancouver, WA 98683-9589 United States
LaHusen, R G (rlahusen@usgs.gov) , USGS Cascades Volcano Observatory, 1300 S.E. Cardinal Court Building 10, Suite 100, Vancouver, WA 98683-9589 United States
Sherrod, D R (dsherrod@usgs.gov) , USGS Cascades Volcano Observatory, 1300 S.E. Cardinal Court Building 10, Suite 100, Vancouver, WA 98683-9589 United States
Iwatsubo, E Y (ewatsu@usgs.gov) , USGS Cascades Volcano Observatory, 1300 S.E. Cardinal Court Building 10, Suite 100, Vancouver, WA 98683-9589 United States
Diefenbach, A (angiedb@usgs.gov) , USGS Cascades Volcano Observatory, 1300 S.E. Cardinal Court Building 10, Suite 100, Vancouver, WA 98683-9589 United States
Thompson, S K (sthompson@usgs.gov) , USGS Cascades Volcano Observatory, 1300 S.E. Cardinal Court Building 10, Suite 100, Vancouver, WA 98683-9589 United States

Following nearly 18 years of eruptive quiescence, a new dacite lava dome began growing in the crater at Mount St. Helens in October 2004. Extrusion was preceded by an intense swarm of shallow earthquakes starting on September 23, by several small explosions starting on October 1, and by remarkable uplift of the south crater floor and glacier. The resulting welt, which was first identified in air photos on September 27, was 450 m wide and 100 m high on October 11 when the first new lava emerged from it. Campaign-style GPS surveys in 2000 and 2003 of a 50-station network concentrated within 10 km of the volcano's summit, but extending more than 30 km radially and covering an area of more than 7400 km2, revealed no surface deformation indicative of magmatic inflation or deflation. A single-frequency continuous GPS station on the 1980-1986 lava dome and annual GPS surveys of points on the dome and surrounding crater floor showed only subsidence of the dome at rates of a few cm/yr, which we attribute to cooling and compaction. A continuous GPS station (JRO1) located 9 km NNW of the vent abruptly started moving toward the volcano, suggesting deflation of a deep magma reservoir, concurrent with the onset of seismicity. Southward motion of JRO1, which is distinctly different from the regional trend of clockwise block rotation in SW Washington, gradually slowed from ~0.5 mm/d before emergence of the new dome to an average of ~0.04 mm/d during the first 11 months of continuous extrusion. Meanwhile, the extrusion rate was relatively steady at ~2 m3/s. Taken together, the JRO1 GPS and extrusion-rate results indicate that the crustal magma reservoir feeding the eruption is being replenished. Twelve new continuous GPS stations were installed on or near the volcano starting in October 2004 by the USGS Cascades Volcano Observatory and UNAVCO Inc., the latter representing the Plate Boundary Observatory. Like JRO1, these stations moved mostly toward the volcano during the ensuing 11 months. A best-fit, spherical, elastic source model constrained by the GPS station displacements from October 2004 to May 2005 indicates a source depth of ~13 km and a net volume change (assuming the bulk modulus of surrounding rock is 30 GPa) of ~16 x 106 m3. The volume extruded, calculated by differencing digital elevation models that span the same time period, is ~45 x 106 m3. This discrepancy suggests that the source has been partially replenished since the eruption began, consistent with the inference drawn from JRO1 and extrusion-rate data. Rapid motions of the growing dome and surrounding crater floor are being measured with helicopter-deployable GPS stations ("spiders"). Lineal extrusion rates during the first few months of the eruption ranged from 0.07-0.12 mm/s (6-11 m/d); subsequent estimates from time-lapse photography were relatively steady at 4-5 m/d. In July 2005, a borehole tiltmeter installed at a depth of 2.3 m on the 1980-1986 lava dome, 440 m NNW of the active vent, began recording reversible tilt episodes with periods of tens to hundreds of seconds, some of which were associated with shallow earthquakes. The tilt signals might indicate recurring stick-slip cycles as a solid plug of dacite lava is forced upward along frictional conduit walls by magma pressure from below (Iverson et al., 2005, this session).

V52B-04

A Dynamical Model of Seismogenic Dome Extrusion, Mount St. Helens, 2004-2005

* Iverson, R M (riverson@usgs.gov) , U.S. Geological Survey, 1300 SE Cardinal Ct. #100, Vancouver, WA 98683 United States

A new mathematical model aims to explain the relationship between two key aspects of the 2004-2005 eruption of Mount St. Helens: (1) for many months the eruption was characterized by nearly constant rates of extrusion (~ 2 m3/s) of a solid dacite plug surfaced with striated fault gouge; and (2) this extrusion was accompanied by repetitive, small (M < 2), shallow (< 1 km) earthquakes that occurred at regular intervals (~ 100 s). The persistence of these nearly periodic earthquakes (dubbed "drumbeats") motivates the hypothesis that they resulted from discrete slip events (< 1 cm each) along the margins of the plug as it was forced incrementally upward by a nearly steady influx of ascending, solidifying magma. This hypothesis is formalized mathematically by considering conservation of mass and momentum of the solid plug and underlying magma, and by adopting simple constitutive relations for the compressibilities of the magma and conduit walls and for the frictional resistance F of the plug sliding against the conduit walls. The resulting model reduces to a nonlinear system of three ordinary differential equations describing simultaneous evolution of the upward plug velocity, u, basal magma pressure, p, and volume of the liquid-filled portion of the conduit, V. In general these equations must be solved numerically, but they also yield analytical results showing that u will persistently oscillate, provided that F does not increase as u increases. Predicted oscillation periods correspond well with the observed interval between drumbeat earthquakes. Moreover, if F decreases nonlinearly as u increases (representing a decay from static to rate-dependent dynamic friction), numerical solutions show that oscillations of u sharpen into discrete stick-slip cycles capable of producing earthquakes. The magma-plug system can be attracted to this type of near-equilibrium, oscillatory behavior even if it begins in a disequilibrium state, as is necessary to instigate a volcanic eruption. However, initial conditions far from equilibrium will cause ejection of the magma plug and a probable transition to explosive behavior.

V52B-05 INVITED

Is the 2004-05 Eruption of Mount St. Helens Tapping New Dacite From the Deep Crust?

* Pallister, J S (jpallist@usgs.gov) , USGS Cascades Volcano Observatory, 1300 Cardinal Court, Suite 100, Vancouver, WA 98683 United States
Thornber, C R (cthornber@usgs.gov) , USGS Cascades Volcano Observatory, 1300 Cardinal Court, Suite 100, Vancouver, WA 98683 United States

The 2004-05 eruption of Mount St. Helens began with seismic activity and uplift of the crater floor in late September 2004, followed by phreatic explosions and extrusion of a lava dome starting on 11 October 2004 and continuing to this time (September, 2005). Since shortly after the first spine of lava appeared, samples have been collected using a steel box dredge (''JAWS '') suspended 60 or 110 feet below a helicopter. This method was developed to acquire samples from the hot and steep-sided dome, which, because of occasional explosions and frequent collapses, has been unsafe to approach on the ground. To date, 21 samples have been collected from the six spines of the new lava dome and petrologic studies of these samples are reported in this session. The lava dome is composed of dacite (65 wt% SiO2) that is geochemically uniform and slightly more evolved than the 1980-86 dacite. The typical lava is crystal-rich with ~50% phenocrysts of plagioclase, amphibole, hypersthene, and Fe-Ti oxides in a microcrystalline matrix that contains ~13% microlites, ~13% glass and ~25% vesicles. Oxide thermometer data for early spine samples cluster at 840-$850 C and NNO+1 log unit. In contrast, samples erupted during the winter of 2004-05 have zoned oxides with apparent temperatures that range to >950 C. Such late-stage heating is likely due to latent heat evolved during rapid groundmass crystallization, or possibly to heating by new magma. Low volatile contents, presence of trydimite and quartz microlites and decreasing H2O with increasing SiO2 in the rhyolite matrix glass indicate extensive shallow (<1 km) crystallization, driven by degassing of water. Major and trace elements of the 1980-86 and 2004-2005 magma batches are similar, which led us to the initial interpretation that the dacite was magma "left over" from the 1980-1986 activity. However, new petrologic and geochemical data suggest instead that the 2004-05 eruption may be fueled by a new batch of dacite magma derived from depth. Geochemically, both Pb and Th isotopes indicate that the 2004 dacite is different from the 1985 dome (Kent et al, this session; Cooper et al., this session). The recent dacite also contains amphibole cores with Al2O3 contents (12-15 wt%) that are too high to be in equilibrium with the MSH dacite at pressures up to 300 MPa (Rutherford and Devine, this session) and suggest incorporation of the amphibole from a higher pressure or from a more mafic magma. Low concentrations of incompatible-elements in MSH dacites indicate derivation of the dacite by melting of lower crustal metabasaltic rocks. Consequently, the isotopic distinctiveness of the 2004-05 dacite and presence of high-Al amphiboles suggest that Mount St. Helens is now being fed by small batches of dacite coming from deeper crustal levels, perhaps in response to unloading of the magmatic system in 1980. These new magmas would have ascended through the remains of the 1980-86-conduit system where they likely mingled with 1980's vintage dacite. The 2004 dacite last equilibrated at P(H2O) of about 140 MPa (Rutherford and Devine, this session), equivalent to depths of about 5 km. Subsequent rise to the surface caused the dacite to degas water and other volatiles, and to crystallize extensively at shallow (<1 km) depths to the point that it erupted as a rheological solid, forming fault-gouge mantled spines. These data have significant implications for the long-term eruptive behavior of Mount St. Helens, as arrival of a new batch of dacitic magma from the deep crust could herald the beginning of a new long-term cycle of eruptive activity.

V52B-06 INVITED

Crystallization of and Conditions in the MSH 2004-05 Dacitic Magma as Indicated by Phenocryst Compositions and Experiments.

* Rutherford, M J (malcolm_rutherford@brown.edu) , Dept. of geological Sciences, Brown University, 324 Brook St.,, Providence, RI 02912 United States
Devine, J D ( ) , Dept. of geological Sciences, Brown University, 324 Brook St.,, Providence, RI 02912 United States

The Mount St. Helens 2004-5 dacitic magma contains plagioclase, amphibole, low-Ca pyroxene phenocrysts that are cyclically zoned in a largely crystalline groundmass rich in plagioclase, pyroxenes, Fe-Ti oxides and glass. The magma also contains small populations of gabbroic xenoliths and some phenocrysts similar to those in 1980-86 magmas; most of these can be identified by their textures. The rim compositions of the phenocrysts are interpreted to reflect pre-eruption conditions in the magma; the magnetite and ilmenite rim compositions yield temperatures ranging from 850C early in the eruption, to 870C and as high as 930 to 950C early in magma erupted early in 2005. Magma samples erupted later in 2005 yield temperatures of 870 C. The complex cyclic zoning in the 2004-5 MSH magma phenocrysts must reflect conditions and events during the pre-erupton history. The zoning in the amphiboles involves a substitution of Fe+Al+Ti for Mg+Si, suggesting a temperature as well as possibly a pressure and composition control. Plagioclase phenocrysts in the samples show a similar number (<3) cycles from An$_{70}$ to An$_{35}$. Interestingly, similar although not identical zonation cycles are present in the MSH magmas erupted in 1986 and earlier, but the zones in the first erupted 1980 magmas are less extreme. The rims of the 1980-86 amphiboles are low in Al2O3 (9-11 wt%) at a given Mg\# compared to the rims of 2004/5 phenocrysts, consistent with the lower temperature of the latter magmas according to compositions of coexisting Fe-Ti oxides. Experiments on crushed samples of the 2004/5 magma show a similar stability field for amphibole as determined for the 1980 samples even though the new magma contains 65 wt% SiO2, ~2% higher than the 1980 magma. However, reversal experiments from 100 to 300 MPa produce large euhedral crystals that are lower in Al than any in the cyclically zoned natural phenocrysts. We conclude that to produce the high-Al amphiboles observed, and the an-rich plagioclase zones, the phenocryst growth had to take place at even greater pressures than 300 MPa, or even more likely, an in-mingling of a mafic magma was involved in the phenocryst growth. In either case, the phenocryst-melt assemblage present just prior to eruption of the 2004/5 magma was no stable at the conditions (850-870C; P(H2O~140 MPa) existing in the magma storage zone, and amphibole would have tended to recrystallize.

V52B-07

Emission Rates, Pre-eruption Gas-Saturation and Ascent Degassing During the 2004-2005 Eruption of Mount St. Helens

* Gerlach, T M (tgerlach@usgs.gov) , U.S. Geological Survey, Cascades Volcano Observatory 1300 SE Cardinal Ct 100, Vancouver, WA 98683 United States
McGee, K A (kenmcgee@usgs.gov) , U.S. Geological Survey, Cascades Volcano Observatory 1300 SE Cardinal Ct 100, Vancouver, WA 98683 United States
Doukas, M P (mdoukas@usgs.gov) , U.S. Geological Survey, Cascades Volcano Observatory 1300 SE Cardinal Ct 100, Vancouver, WA 98683 United States

Intermittent airborne measurements of volcanic gases began on 27 September 2004 during the initial unrest. Target gases included CO2, SO2 and H2S. The average emission rates through the end of July 2005 were approximately 650 metric tons/day (t/d) CO2 and 100 t/d SO2; H2S was always <8 t/d. The cumulative emissions are estimated to be about 190 kt CO2 and 30 kt SO2. The emission rates are significantly lower than those of 1980 and the earlier dome-building episodes of the 1980s, which frequently produced several thousand t/d CO2 and several hundred t/d SO2. Nevertheless, the average CO2/SO2 of 10 (molar) for the present emissions is similar to that of the 1980-81 emissions. Despite the lower emission rates, the cumulative CO2 released and the estimated amount of melt supplied from depth (46 Mt) indicate the dacite would have been gas-saturated at plausible chamber depths. Data on the S content of melt inclusions and matrix glasses underestimate observed SO2 emissions by over a factor of five, corroborating pre-eruption gas saturation of the dacite. Using the 1980 magma equilibration depth (8 km) for comparison and assuming a similar water content (5 wt%) for the rhyolitic melt fraction (Rutherford, 1993; Blundy and Cashman, 2001) give a gas fraction constrained by H2O-CO2 solubility data of 1-2 vol% for the dacite. This compares with an estimate of 10-15 vol% for the 1980 dacite prior to eruption and of 5-30 vol% for common intermediate to silicic magmas (Wallace, 2003). Closed system ascent degassing calculations show that the volume fraction of gas increases to 8 vol% at 4-5 km and reaches 50 vol% at 1 km, where final solidification begins. The gas fraction can potentially increase to >60 vol% during solidification. Allowing for gas separation during extrusion, these results are consistent with observed dacite vesicle fractions averaging 25 vol% (Pallister et al. this session). Ascent degassing calculations also predict melt water contents similar to values measured on rare glassy dacite fragments last equilibrated at depths of 1.2-1.8 km (Mandeville this session).

V52B-08

Time Series 210Pb-210Pb Data for Lavas Erupted From Mount St. Helens Volcano: Implications for Time-scales of Degassing and Crystallization

* Reagan, M K (mark-reagan@uiowa.edu) , University of Iowa, Dept. of Geoscience, Iowa City, IA 52242 United States

210Po has been shown to be highly volatile at magmatic temperatures and generally degasses almost completely during eruption. This has been true for nearly all of the lavas erupted from Mount St. Helens since November 2004, as the last day of complete degassing of 210Po from the Mount St. Helens dacite generally corresponds with their day of eruption within error of analysis and knowledge of eruption day. In contrast, samples SH 304 and SH 305, which erupted in October and November 2004, apparently last degassed Po about the day of the first phreatic explosions. Our sample of a lithified, interior portion of the gouge coating on a dacite spine erupted in April, 2005 also degassed Po until it erupted. The exterior less-lithified gouge erupted about July 1, 2005, however, was strongly enriched in 210Po over 210Pb, indicating that it was a primary conduit for escaping magmatic volatiles. Comparing our results with unpublished 226Ra activities (Donnelly and Cooper, this session) show SH 304 and SH 305 whole rocks to have near equilibrium (210Pb/226Ra) values. If the significant enrichment in Li observed for plagioclase in these samples by Adam Kent (personal communication, 2005) was generated by a flux of a Li and Rn-bearing volatile phase, then the duration of this fluxing must have been over a period that was long enough for Li to equilibrate in plagioclase, but too short to cause a significant 210Pb excess from 222Rn decay (i.e. less than 1-2 years). A preliminary 210Pb/226Ra ratio measured for plagioclase extracted from SH 304 yielded a model age of several decades, showing that a significant portion of the plagioclase grew in the decades leading to eruption. However, the lack of 210Po/210Pb disequilibrium in this same separate suggests that little of this growth occurred during the last year before eruption.